Trait RLComponentsTypes 

pub trait RLComponentsTypes {
    type Backend: AutodiffBackend;
    type Env: Environment<State = Self::State, Action = Self::Action> + 'static;
    type EnvInit: EnvironmentInit<Self::Env> + Send + 'static;
    type State: Into<<Self::Policy as Policy<Self::Backend>>::Observation> + Clone + Send + 'static;
    type Action: From<<Self::Policy as Policy<Self::Backend>>::Action> + Into<<Self::Policy as Policy<Self::Backend>>::Action> + Clone + Send + 'static;
    type Policy: Policy<Self::Backend, Observation = Self::PolicyObs, ActionDistribution = Self::PolicyAD, Action = Self::PolicyAction, ActionContext = Self::ActionContext, PolicyState = Self::PolicyState> + Send + 'static;
    type PolicyObs: Clone + Send + Batchable + 'static;
    type PolicyAD: Clone + Send + Batchable;
    type PolicyAction: Clone + Send + Batchable;
    type ActionContext: ItemLazy + Clone + Send + 'static;
    type PolicyState: Clone + Send + PolicyState<Self::Backend> + 'static;
    type LearningAgent: PolicyLearner<Self::Backend, TrainContext = Self::TrainingOutput, InnerPolicy = Self::Policy> + Send + 'static;
    type TrainingOutput: ItemLazy + Clone + Send;
}

All components used by the reinforcement learning paradigm, grouped in one trait.

Required Associated Types

type Backend: AutodiffBackend

The backend used for training.

type Env: Environment<State = Self::State, Action = Self::Action> + 'static

The learning environment.

type EnvInit: EnvironmentInit<Self::Env> + Send + 'static

Specifies how to initialize the environment.

type State: Into<<Self::Policy as Policy<Self::Backend>>::Observation> + Clone + Send + 'static

The type of the environment state.

type Action: From<<Self::Policy as Policy<Self::Backend>>::Action> + Into<<Self::Policy as Policy<Self::Backend>>::Action> + Clone + Send + 'static

The type of the environment action.

type Policy: Policy<Self::Backend, Observation = Self::PolicyObs, ActionDistribution = Self::PolicyAD, Action = Self::PolicyAction, ActionContext = Self::ActionContext, PolicyState = Self::PolicyState> + Send + 'static

The policy used to take actions in the environment.

type PolicyObs: Clone + Send + Batchable + 'static

The policy’s observation type.

type PolicyAD: Clone + Send + Batchable

The policy’s action distribution type.

type PolicyAction: Clone + Send + Batchable

The policy’s action type.

type ActionContext: ItemLazy + Clone + Send + 'static

Additional data as context for an agent’s action.

type PolicyState: Clone + Send + PolicyState<Self::Backend> + 'static

The state of the parameterized policy.

type LearningAgent: PolicyLearner<Self::Backend, TrainContext = Self::TrainingOutput, InnerPolicy = Self::Policy> + Send + 'static

The learning agent.

type TrainingOutput: ItemLazy + Clone + Send

The output data of a training step.

Implementors

impl<B, E, EI, A> RLComponentsTypes for RLComponentsMarker<B, E, EI, A>

where B: AutodiffBackend, E: Environment + 'static, EI: EnvironmentInit<E> + Send + 'static, A: PolicyLearner<B> + Send + 'static, <A as PolicyLearner<B>>::TrainContext: ItemLazy + Clone + Send, <A as PolicyLearner<B>>::InnerPolicy: Policy<B> + Send, <<A as PolicyLearner<B>>::InnerPolicy as Policy<B>>::Observation: Batchable + Clone + Send, <<A as PolicyLearner<B>>::InnerPolicy as Policy<B>>::ActionDistribution: Batchable + Clone + Send, <<A as PolicyLearner<B>>::InnerPolicy as Policy<B>>::Action: Batchable + Clone + Send, <<A as PolicyLearner<B>>::InnerPolicy as Policy<B>>::ActionContext: ItemLazy + Clone + Send + 'static, <<A as PolicyLearner<B>>::InnerPolicy as Policy<B>>::PolicyState: Clone + Send, <E as Environment>::State: Into<<<A as PolicyLearner<B>>::InnerPolicy as Policy<B>>::Observation> + Clone + Send + 'static, <E as Environment>::Action: From<<<A as PolicyLearner<B>>::InnerPolicy as Policy<B>>::Action> + Into<<<A as PolicyLearner<B>>::InnerPolicy as Policy<B>>::Action> + Clone + Send + 'static,

type Backend = B

type Env = E

type EnvInit = EI

type LearningAgent = A

type Policy = <A as PolicyLearner<B>>::InnerPolicy

type PolicyObs = <<A as PolicyLearner<B>>::InnerPolicy as Policy<B>>::Observation

type PolicyAD = <<A as PolicyLearner<B>>::InnerPolicy as Policy<B>>::ActionDistribution

type PolicyAction = <<A as PolicyLearner<B>>::InnerPolicy as Policy<B>>::Action

type ActionContext = <<A as PolicyLearner<B>>::InnerPolicy as Policy<B>>::ActionContext

type PolicyState = <<A as PolicyLearner<B>>::InnerPolicy as Policy<B>>::PolicyState

type TrainingOutput = <A as PolicyLearner<B>>::TrainContext

type State = <E as Environment>::State

type Action = <E as Environment>::Action